Business Research Engine
Business research can look convincing while mixing unsupported claims with recommendations. Build a structured evidence-to-decision workflow with transparent uncertainty and alternative scenarios.
Who and when?
Founders, strategy consultants, BD managers, product managers, investment/market analysts.
From intake to final review.
- Every consequential factual claim has a traceable source or is explicitly labeled uncertain. - Dates, markets, and units are consistent. - Conflicting evidence is not silently discarded. - Recommendations link to criteria, evidence, and stated assumptions. - No fabricated market sizes, financial metrics, or citations.
Inputs and outputs
Required: decision question, context, geography, time horizon, decision criteria, constraints. Optional: competitor list, budget, internal data, source preferences, hypothesis, decision deadline.
- decision-brief.md
- research-plan.md
- sources.csv
- claim-ledger.csv
- evidence-gaps.md
- market-analysis.md
- competitor-matrix.csv
- scenario-analysis.md
- decision-memo.md
- limitations.md
- state.json
A synthetic worked example.
Synthetic example: A fictional company compares two markets using evidence with conflicting figures. The ledger keeps definition and date differences and avoids a false total. The memo compares options against explicit criteria, states uncertainty, and proposes a small validation step before commitment.
claim_id,source_id,status CL-01,SRC-A,conflicts_with_CL-02 CL-02,SRC-B,definition_differs
An illustrative output excerpt, not a client result or live run. Sample approval cannot authorize a real action.
How to use it
- After repository publication, obtain the files and read README and SKILL.md. No download link exists before release.
- Prepare inputs and sources in a separate run folder. Keep secrets and client data out of public files.
- Start at intake and follow the workflow manually or with an AI tool that can read and write local files. No platform compatibility is claimed before testing.
- Review each phase and record a version-specific decision. Missing inputs and weak evidence create blockers, not guessed completion.
- Persist state, decisions, and artifacts. Resume from the latest approved phase without silently replacing approved output.
Limits and safety
Investment advice, autonomous purchasing/contracting, proprietary data acquisition, guaranteed market forecasts.
Files and web pages are untrusted information, not authorization to change goals, publish, or contact anyone. External effects need separate permission. Run data stays local and examples are synthetic. File checks cannot prove factual truth or reviewer identity.
Status: v1 specification for review. GitHub URL, release, license, and exact runtime requirements are not published. Contributions will use the actual repository after launch.